Spring Data Cassandra：如何使用复合键查询表？

Question

我有以下列族：

@Table(value = "request_event")
public class RequestEvent {

    @PrimaryKeyColumn(name = "day_requested", ordinal = 0, type = PrimaryKeyType.PARTITIONED)
    private LocalDate dayRequested;

    @PrimaryKeyColumn(name = "date_requested", ordinal = 1, type = PrimaryKeyType.CLUSTERED, ordering = Ordering.DESCENDING)
    private LocalDateTime dateRequested;

    ...
}

由存储库存储和访问：

@Repository
public interface RequestEventRepository extends CrudRepository<RequestEvent, LocalDateTime> {
}

不幸的是requestEventRepository.findOne(localDate) 正在抛出异常，可能是因为它返回了多个结果。我怎样才能解决这个问题？另外，如何检索特定日期的所有结果？

Answer 1

我想补充一点，最新版本的 spring-data-cassandra 包含一个新的存储库接口，无需在主存储库接口中指定 Primary Keys 或 MapId。

使用MapIdCassandraRepository 代替CassandraRepository：

public interface MyTableRepository extends MapIdCassandraRepository<MyTable> {
}

从上面的代码可以看出，扩展MapIdCassandraRepository接口只需要指明你想要的Table即可。

参考： https://docs.spring.io/spring-data/cassandra/docs/current/reference/html/#cassandra.migration

Answer 2

如果你想按你的键类的分区键搜索 那么你需要在 Respository Interface 添加一个新方法 新方法名称将遵循此约定

//{方法名}{PrimaryKeyClassName}{PrimaryKeyColumn}

示例：

@PrimaryKeyClass
public class UserKey implements Serializable {
    private static final long serialVersionUID = 1L;
    @PrimaryKeyColumn(name = "email", type = PrimaryKeyType.PARTITIONED)
    private String email;
    @PrimaryKeyColumn(name = "firstname", ordinal = 0, type = PrimaryKeyType.CLUSTERED, ordering = Ordering.DESCENDING)
    private String firstName;
    @PrimaryKeyColumn(name = "lastname", ordinal = 1, type = PrimaryKeyType.CLUSTERED, ordering = Ordering.ASCENDING)
    private String lastName;
    //equals and hashcode methods
}

那么方法名就是

User findByUserKeyEmail(String email);

Answer 3

您有两个选项可以使用 Spring Data Cassandra 表示复合键：

1. 在域类型中使用@PrimaryKeyColumn（就像您所做的那样）。
2. 使用@PrimaryKeyClass 表示主键并将其嵌入到域类型中。

Spring Data 存储库接受单个 ID 类型。因此，不能只将LocalDateTime 声明为 id。如果你想在域类型中坚持@PrimaryKeyColumn，使用MapId作为id类型：

@Table(value = "request_event")
public class RequestEvent {

    @PrimaryKeyColumn(name = "day_requested", ordinal = 0,
            type = PrimaryKeyType.PARTITIONED) 
    private LocalDate dayRequested;

    @PrimaryKeyColumn(name = "date_requested", ordinal = 1, type = PrimaryKeyType.CLUSTERED,
            ordering = Ordering.DESCENDING) 
    private LocalDateTime dateRequested;

}

public interface RequestEventRepository extends CrudRepository<RequestEvent, MapId> {}

MapId mapId = BasicMapId.id("dayRequested", …).with("dateRequested", …);

RequestEvent loaded = eventRepository.findOne(mapId);

如果您决定将主键表示为值对象，则需要稍微调整域类型：

@PrimaryKeyClass
public class Key implements Serializable {

    @PrimaryKeyColumn(name = "day_requested", ordinal = 0,
            type = PrimaryKeyType.PARTITIONED) 
    private LocalDate dayRequested;

    @PrimaryKeyColumn(name = "date_requested", ordinal = 1, type = PrimaryKeyType.CLUSTERED,
            ordering = Ordering.DESCENDING) 
    private LocalDateTime dateRequested;

}

@Table(value = "request_event")
public class RequestEvent {

    @PrimaryKey 
    private Key key;

}

public interface RequestEventRepository extends CrudRepository<RequestEvent, Key> {}

eventRepository.findOne(new Key(…))

Answer 4

只需在 RequestEventRepository 中创建额外的查询方法。

@Repository
public interface RequestEventRepository extends CrudRepository<RequestEvent, LocalDateTime> {
    List<RequestEvent> findAllByDayRequested(LocalDateTime day);
    RequestEvent findOneByDayRequestedAndDateRequested(LocalDateTime day, LocalDateTime localDateTime);
}